DCC Explained

[Roger Sunderland]

While I appreciate some of the detail here, look at what the OP has asked and think about their level of skill.I've been modelling for a couple of years now and I still don't understand much about wiring, resistance levels, ohms, wattage, blah blah blah.OP: Look online at videos and ask people for specifics to what they use for wires.Don't get bogged down with the detail unless it interests you. I bought wires, LEDs and power supplies that others have and it's served me well - without knowing all the resistance details, or what wires are what size or any of that stuff.I have "bus wires", "dropper wires" and "chip wires" <-- three different sizes depending on what I need. That's my sort of level of technical skill compared with 7/0.2 thickness wire.As I say, it's great going into detail but I don't think the OP will have a clue what some people are writing.

 

Neither will 99.9% of everybody else!

[newbryford]

Oh dear

How did these questions:

end up in a definition war!

 

Keith

 

Agreed.

As a confirmed DCC user, I can completely sympathise with the sentiment of post#22

Edited September 14, 2018 by newbryford

[Mallard60022]

Oh dear

How did these questions:

end up in a definition war!

 

Keith

I think most of us know the answer to that.......................enough to put the poor poster off for life.

How I wish people were a weeny bit less terratorial and pedantic about stuff.............

Phil

[WIMorrison]

As a young engineer I was taught that the really good engineer knows how to answer a question with just enough information to inform the questioner, without providing excess information that detracts from the value of the response and only confuses the questioner.

 

The really good engineer keeps the intricate detail for those who will understand and can use such detail and will make it available at the appropriate time and to the correct audience.

 

A mantra that has served me will over the many years that I have used it.

[blueeighties]

Oh Dear - Junction Mad wrote: "A few things ...," which need a correcting or clarifying response to avoid future confusion or problems

"While the ability of the wire to carry the full rated DCC short circuit current is good advice , it is rather overkill and can mean in theory, the droppers become impractically large."

Phil: '(Over)KILL is what I wanted people to Avoid ! - Seriously !  YOUR WIRING MUST MATCH the capability of the the SOURCE ( here the Power Supply to,  and output from, the DCC Central Controller OR Booster if more than 1 'Power District' (NMRA Terminology).

 

FAULTS WILL OCCUR (despite what some people appear to write) - and the (fast) internal protection of the Central Controller is designed to protect both the User AND the equipment in that case: by stopping the current flowing ONCE THE EXCESS HAS BEEN DETECTED ....  This is why I drew attention to the importance of realising that this is related to YOUR CHOICE of Controller and is also of its MAXIMUM NORMAL** (Continuous maximum) Current output Rating:  For some controllers this may be only 1A, but for others it might easily be 3,4,5,8, 10 or 12 Amps. {**This has nothing to do with how much current your layout is usually taking }

 

YOUR WIRING (and your responsibility) MUST ENSURE that this FULL MAXIMUM current can flow CONTINUOUSLY to ALL PARTS of your LAYOUT [ EXCEPT beyond  Local sub-district breakers giving a lower local limitation: THIS IS THE EQUIVALENT of the Mains Electricity reaching your house [100A continuous capability]  and then passing through Fuses/RCCBs with 5A, 15A 30A etc breakers for the local sub-district circuits beyond (eg lighting, ring main, cooker)  AND ALSO for the 30A Ring Main with 13A Sockets, via Plugs (3A-13A fuses) to devices which then have smaller protective fuses inside them .... thinner wire being acceptable in each of the later stages providing it can take the  (Protected) MAXIMUM Rated current CONTINUOUSLY.   

Note that an old fashioned fuse is a thin piece of wire - designed to melt in an enclosure which will not catch fire.  A 'Dropper' ( or should that be 'riser' wire 8-) )  is short, and is connected to adjacent metal which helps conduct heat away from it - its resistance is. or should be 'negligible' - especially in comparison to the TOTAL RESISTANCE of the  Electrical LOOP from the controller via busses  to ANY PART of the track.   FOR THE SHORT_CIRCUIT PROTECTION to WORK - the time taken is irrelevant to this calculation -  V=IR.  Simple Ohms Law:   

Example: If the NORMAL MAXIMUM CURRENT is 1A, and the TRACK VOLTAGE  12V: From V=IR ...  R= V/I  ...   12/1 = 12ohms BUT

IF THE TRIP CURRENT is then 2A (just 1 A or 100% over 'normal maximum')  then from V=IR...  Rmax = V/I ...  12/2    <  6 ohms MAX

 

BUT if a larger output controller is chosen -  eg 5 Amp Normal Maximum Output:  From V=IR ... R= V/I ... 12/5  =    2.4 ohms BUT for the

SHORT CIRCUIT DETECTION -  requiring (even a MOMENTARY current of )  say 7 A or 10 ..:  R= V/I ... 12/7 or 12/10 ohms < 1.2ohms

As a check: My garden line (125m total) has a loop resistance of 0.5 ohms .  20V = I x 0.5 .. would allow a 40A short circuit current to flow  [ Note also, that the slightly higher track voltage allows for more resistance in the track, and motor currents are proportionally lower ]

My 00/H0,H0e and G-scale portable layouts all use 16Vdcc [ by which I mean 0-peak measured by 'scope or Rrampmeter etc ]

 

For 4mm scales the largest practical size is around 7/0.2 , for O 16/0.2.

I do not mention any size of wire, particularly for the final connection to the track - as the scale is not relevant to the electrons heating up your copper wire as they flow around.

However, as a GUIDE, the quoted ratings on wires sold for 230Vac use are based on the heating effect/dissipation when embedded in a plaster wall or other appropriate environment -  For my loft layout I use '3A' or '6A' twin flex (using the coloured outer sleeves for identification)  - PROVIDED the circuit passes the 'coin test' as calculated above -- AND ALSO checking for less than a couple of volts drop in any section when  'the maximum load'  is present in the loop --- decoders might not work below 7V (many current designs do -  this is not usually a problem if the 'coin test' has been passed).

 

 

Most DCC short circuit limiters work very fast , hence the wire has only to carry the fault current for a few milliseconds , hence these wires are suitable . However the DCC “ bus” wires should as large as practical , I use 2.5sqmm cable

Phil: ALTHOUGH - 'heatwise' they SHOULD only be carrying the EXCESSIVE current for a few millseconds - if the LOOP RESISTANCE is TOO HIGH then the TRIP VALUE will NEVER be reached ....and heating will continue - revealing itself first at the weakest part 

{My experience of this was a failed internal switch inside a Fleischmann point - it was taking a continuous 3.5A - and therefore NOT tripping the ZTC511 Controller with a 5A normal-maximum continuous output capability - result - 1 melted point} 

 

However if you have a large layout or for whatever reason a large capacity DCC system, you need to consider breaking up the layout into DCC “power districts “ and installing DCC cutoffs locally in each power district, this keeps the individual power districts short circuit currents within manageable levels.

Phil: Agreed except for the 'However' - the SUB-DISTRICTS (if from the same power supply) or POWER-DISTRICTS ( own power supply ie Controller and Boosters)  are desirable to assist FAULT_FINDING by Localising the problem: 'divide and conquer' - and are best chosen for that purpose, and not  for historical 'cab-control' sectionalising reasons. In the same way your house has a 'consumer unit' with a range of circuit breakers as described earlier, I strongly recommend the use of PSX or similar intelligent circuit breakers because it permits 'more modest' wiring  BEYOND the protective device  BUT especially once Sound and lighting is used as these are designed to cope with their surges ( in the same way as there are Type 1,2,..3  circuit breakers available)

 

Note that DCC is more correctly called a DC pulse train , it may or may not have alternating feeds , if it has it’s an AC pulse train .

The voltage is specified as peak to peak and the NMRA publish reccomended voltages for each gauge , this is 15v for HO/ OO, ( that’s not 20 v peak, just 15 )

Phil: Where does this mish-mash of confusion and error come from?  

IT IS NOT HELPED by a COMMON ERRONEOUS ASSUMPTION that 'dc' means a smooth steady single value supply like from a battery, or that 'ac' means a Sinusoidal waveform like the Mains Electricity supply:  these are both just specific EXAMPLES of dc & ac.

There are many situations where CONTEXT is the best selector of the terminology quoted- and it usually depends on HOW YOU PROCESS the SIGNAL  - as 'analogue' or 'digital' for example.

 

DC - Direct Current  = Flowing in one direction only.  It may vary in value - slowly or quickly, turned on or off or even get reversed in polarity (direction of flow) in some devices - CONTEXT is used to select some aspects of the terminology to best  relate to the environment OR PROCESSING in which it is encountered. DC 'can be considered' as AC with a frequency of 0 Hertz if more convenient.

 

Simply put: BLOCKED IN ONE DIRECTION by a DIODE ('semi-conductor) - BLOCKED (after charging up),a CAPACITOR,but passed through un-impeded by an INDUCTOR.  The type of input required by a motor-with commutator for continuous rotation to take place.

 

AC - Alternating Current = Flowing in both directions - A Diode in series will remove the reversed-part of the current,  and a capacitor across the supply will be seen as a short-circuit  (which is why the interference suppresion capacitors NEED to be REMOVED for dcc)

 

So a Battery Voltage is 'dc' - pure and simple.   Sound pressure waves are 'AC' ( Barometric Pressure is 'DC' ) A loudspeaker coil responds ONLY to the AC content of the 'sound' waveform applied and an old analogue TV/video waveform is 'DC' although it varies from (below black) to white. -   'DC restoration' (clamping) was required to maintain black  as black on a TV when the average value varied.

 

A Modern Electronic Controller - such as that from Gaugemaster or Roco, produces a 'dc' output which is a series of pulses of varying width and/or amplitude - to apply to a motor with a commutator (that requres a 'dc' supply.) - a diode could be placed in series,and the motor would still be able to run in that direction  - but not the other way.

An old H&M Duette produces a rectfied-sine wave shape waveform - peaking at 28V 100x a second - but is labelled 12Vdc by H&M..

Under load, at  'maximum speed' the average voltage is 12V, and, for any one direction of the control  knob it is DC - one direction - but certainly not smooth  because it ALSO HAS a (large) AC component.

 

A DCC WAVEFORM IS AC - it is a series of pulses at a frequency a bit below 10,000 Hz -  alternate pulses cause current to flow in opposite directions. THE AVERAGE VALUE  is also ZERO   ( ie NO 'DC COMPONENT) - except when/if a 'Loco 0' unfitted analogue loco is being driven - but this is not recommended  and not allowed on many systems.  

 

It is a 'square wave' in that it has fast-rising edges whose timing indicates the data it s carrying - although,as with ALL devices, the rise is controlled to  minimise interference etc.    It is 'digital' in that the duration of the pulses is measured by the receiving decoder, and classified as either a    '1' or a '0' .  In the Decoder, the track AC signal is Full-wave rectified to produce a DC power supply for the decoder to operate.  From this internal supply,  Function outputs and the MOTOR can be controlled - as stated by PWM for efficient variable speed  - and this is commonly via an 'H-bridge - whereby swapping which pair of ( 4 )  transistors are used, the direction can be selected.  However,decoders are also available to drive Servos and AC motors - this is NOT a requirement.  Many european decoders are also multi-protocol and multi-standard for motor types.

 

THE VOLTAGE may be measured and described EITHER as a 0-peak value   eg 12V   16V up to a maximum of 22V,  or, as is common on an oscilloscope; as peak-peak which would be double these figures.   There is no Fixed DCC track voltage specified in the NMRA standard  - just recommendations and a maximum limit (23V) which keeps it on the safe side of certain regulations. Prior to modern electronic controllers and switched-mode power supplies, Motor/Track Voltages varied considerably - eg Tri-ang controllers stating 0-14V BUT also being higher off-load due to lack-of voltage regulation.    RocoDCC default outputs 16V (dcc0-peak) which gives a maximum of about 12V at the motor. 

 

The control-data of DCC is carried by the MODULATON of the voltage using 'FM'/ Bi-phase-Mark or Manchester1 or other names which can describe the idea of a 0 and 1 being sent by pulses of (2) different durations - 1 half the other  (Except during 'Loco0'mode )

 

[HOW a waveform is BEST described usually depends on HOW YOU intend PROCESSING IT !! - For example, 'Digital TV Transmissions'  are transmitted by a process which has some very analogue characteristics !!! - which can be decoded or displayed as a 'grid' of 16 x 16 dots .. "256 values"' - not just 2 binary levels ....and therefore, if you have too much amplification between your aerial and TV receiver - such that old-fashioned 'clipping' occurs - then your TV will show good signal strength BUT low quality - and be unable to decode the picture 8-)]   { And different ,TV multiplexes use different settings - so reception of some is more difficult than others )

 

A multimeter will give you an inaccurate voltage reading on DCC , what it will read will be a function of the individual meter

Phil: Agreed with many 'cheap' digital meters which are only calibrated for Sne Waves, and a t 50-60Hz. eg Flukes are accurate on DCC.

Note that most DCC decoders don’t produce DC to the motor from the DCC signal , what they typically do is rectify the DCC , to DC , then chop that DC into a PWM ( pulse width modulated ) AC signal and feed that to the motor

Phil: Earlier the writer wanted to call the AC DCC waveform DC, and now appears to be saying the feed to a dc motor is 'ac' !!!

As described above - PWM chopping DC as in most decoders results in a DC waveform to the motor - many now operate at ultrasonic frequencies ( early decoders continued the 50-60Hz (100/120Hz) mains relationship as with Zero-1 - producing noticeable 'jerks' n the motor, and unsuited to coreless motors ) - this has now allowed smaller components to be used - saving space. The motor's inductance helps smooth out the pulses.   In between the 'digital' pulses, as the motor continues to spin,it acts as a generator - creating a voltage - a back emf - and this can be measured by the decoder to trim the speed - this is an example of analogue measurement on-board by an A to D converter

 

I never suggest a particular size of wiring - as the real question is does it meet the NECESSARY REQUIREMENT - and the first requirement is your continued safety - and simply checked with a 'coin test' short circuit - does it protect ???  THIS TEST can be done on a reel of cable .... if the 100m reel passes the test - then any part of it will meet the test.  However - don't forget the resistance of Nckel Silver Rail - especially in smaller gauges / 'code' eg 55 or 80 in N,  75,80,100,120 in 00/H0 332 in G and in brass - a better conductor!!

Christ on a bike...I though our hobby was meant to be fun!! Sounds as though you are worried about catastrophic explosions.....

[Roger Sunderland]

Christ on a bike...I though our hobby was meant to be fun!! Sounds as though you are worried about catastrophic explosions.....

It is, it really is! Others may know, but, irrespective of incomprehensible content, is this the longest post in the history of BRM?

[brianusa]

All this has reminded me why I returned to tinplate trains.  Not that I begrudge those who enjoy such complications and no doubt their layouts run smoothly and give pleasure and obviously most seem to know what they are doing.  However, I am reminded of the instruction catalogue (its no longer a leaflet!) of things to do that comes with a TV set these days or your printer or even your telephone, mobile or otherwise.  Where did simplicity go and why did so many complications take its place?

 

Brian.

[Benbow]

Strewth if I had known Railway Modelling was that complicated I would have taken up knitting!

[melmerby]

All this has reminded me why I returned to tinplate trains.  Not that I begrudge those who enjoy such complications and no doubt their layouts run smoothly and give pleasure and obviously most seem to know what they are doing.  However, I am reminded of the instruction catalogue (its no longer a leaflet!) of things to do that comes with a TV set these days or your printer or even your telephone, mobile or otherwise.  Where did simplicity go and why did so many complications take its place?

 

Brian.

When did you last buy something electrical?

Most seem to come with online catalogues as a print version would need a whole forest just to supply the paper!

 

Keith

[brianusa]

Of course you are right, Kevin.  Probably this occurred with our first Smart TV, the one before the current updated example with newer and newer so called benefits that aren't much better than the old one, which did come with a mighty instruction book that I didn't delve far into.

     Incidentally, our new car came with a 600+ page instruction book; barely fits the glove box!  Glove box - now there an anachronism; who puts gloves in there anymore!  I forget what they're called back home!  Glove box?

 

Brian.

[Oakydoke]

Others may know, but, irrespective of incomprehensible content, is this the longest post in the history of BRM?

No, I think there have been longer ones....from the same source.

[Andymsa]

Ask 50 people the same question your get 50 different view points.

 

That long post is a good remedy for insomnia.

[Junctionmad]

The poor old OP must be even more confused, it's like reading war and piece. There are good points to all the posts but all he wanted was basic advice On what is required.

So I just say how I did mine, there are six boosters that each are subdivided, single multi strand 2.5mm mains wire is used these go to bdl168s the feeds from these are 16/02 these are connected to the rail by 0.7 copper wire in the visible sections and direct in non visable sections. All cables are colour coded

+1 , as I said your selection is very common in my experience

I hope you tested that 0.7 wire in 100 meter lengths for short circuit survability

Edited September 15, 2018 by Junctionmad

[Right Away]

Most informative thread, helps me make my mind up, I shall stay well away from DCC and remain analogue.

Appreciated, the more in-depth technical information posted in this thread can be interesting to some, with quite viable points for owners of very large layouts but providing sensible electrical practices are followed, DCC need not be viewed with trepidation.

 

I have not had to resort to utilising my electrical calculations/AC theory from "Tech" days in order to build my garage-sized 4mm scale DCC railway, which has operated well for years. A good multimeter is a wise investment for fault finding.

 

If anything, the simple expedient of running 4mm mains cable (I know, overkill) and feeding EVERY piece of trackwork with 7/0.2mm dropper wires is far less hassle than providing independently switched isolated sections.

 

Initial outlay for the necessary decoders and control system is the only downside.

[GoingUnderground]

I think the OP should go and buy a few books on DCC, such as Ian Morton's "Digital Command Control" now in its 2nd edition, in the "Aspects of Modelling" series, and read them at his leisure. DCC isn't something you can just jump in to without some understanding of the basics. I read this book and it helped me get started, along with Alan Gartner's Wiring for DCC website, and latterly Brian Lambert's site. But I found them more useful after reading the book as I found that I then understood better the content of their websites.

Edited September 16, 2018 by GoingUnderground

[Andymsa]

+1 , as I said your selection is very common in my experience

I hope you tested that 0.7 wire in 100 meter lengths for short circuit survability

The 0.7 wire is solid non flexible copper wire and is only 1 inch long

[VicZA]

Folks .. again ... thanks for all the info so far - greatly appreciated.

 

A few more questions if I may ...

 

I have just received my latest loco which is a previously owned 8750 Tand (a beauty!) but I think it is only DC, so:

(a) how do you tell if it is DC, DCC or Analog bu looking at it "under the hood"?

(b) assuming it is DC - which I think it is - can I run it on a DCC track using a DCC controller (I have the Hornby R8213)?

© If I want to change it to DCC, what decoder do I need?

(d) Does anyone have a link to a you tube video (or similar) to show how this is done.

 

Thanks

[newbryford]

Folks .. again ... thanks for all the info so far - greatly appreciated.

 

A few more questions if I may ...

 

I have just received my latest loco which is a previously owned 8750 Tand (a beauty!) but I think it is only DC, so:

(a) how do you tell if it is DC, DCC or Analog bu looking at it "under the hood"?

(b) assuming it is DC - which I think it is - can I run it on a DCC track using a DCC controller (I have the Hornby R8213)?

© If I want to change it to DCC, what decoder do I need?

(d) Does anyone have a link to a you tube video (or similar) to show how this is done.

 

Thanks

 

A catalogue number search may reveal if it is DCC ready or not. (I have an early version of the 8750 and it has no socket whatsoever and I haven't tried to fit DCC to it yet.) Otherwise you may have to look inside to see if it has a DCC socket or not - this will determine what you need to do to make it run on DCC and that will probably determine which decoder is required.

 

Some DCC controllers support running one DC loco - usually as address 0, but this is not generally a good idea as the DCC signal will gradually fry the motor...…..

 

There a quite a few DCC fitting guides out there - just type in your loco and accompany it with "DCC fitting" should throw up a few .

 

Cheers,

Mick

[Nigelcliffe]

Folks .. again ... thanks for all the info so far - greatly appreciated.

 

A few more questions if I may ...

 

I have just received my latest loco which is a previously owned 8750 Tand (a beauty!) but I think it is only DC, so:

(a) how do you tell if it is DC, DCC or Analog bu looking at it "under the hood"?

you could look, which is definitive. Or put it on a programming track and try to read it. If nothing, then its either analogue, or has a rubbish decoder fitted.

 

 

(b) assuming it is DC - which I think it is - can I run it on a DCC track using a DCC controller (I have the Hornby R8213)?

A few DCC controllers have the ability to run an analogue loco on "address zero". This is a BAD THING. Its in the old specifications as a way to get people to change, in the 1990's, from ananalogue to DCC, but its horrible to motors, subjecting them to lots of AC power. The motors buzz like crazy, overheat, and can burn out if subjected to this abuse.

 

So, don't do it, even if the Hornby controller will let it happen.

 

 

© If I want to change it to DCC, what decoder do I need?

One that fits the available space, and with enough power to run the loco.

 

Suggest starting with quality decoders from the outset, so Lenz Basic, or one from Zimo. Don't waste time with inferior ones.

 

As to what fits, and how much work fit it, that depends on the vintage of the model. Is it really a 1970's model, or something more recent ?

 

(d) Does anyone have a link to a you tube video (or similar) to show how this is done.

See answer above from Mick.

 

 

[RAF96]

@post #44

Quote - So, don't do it, even if the Hornby controller will let it happen.

 

The R8213 Select controller is set default at the latest version (1.5) to run a loco zero, but as stated it is not recommended.

 

@OP

 

Is it DC or DCC? - as stated by others stick it on your DCC track (briefly) and if it makes a racket standing still it is DC.

 

Rob

Edited September 18, 2018 by RAFHAAA96

[melmerby]

Folks .. again ... thanks for all the info so far - greatly appreciated.

 

A few more questions if I may ...

 

I have just received my latest loco which is a previously owned 8750 Tand (a beauty!) but I think it is only DC, so:

(a) how do you tell if it is DC, DCC or Analog bu looking at it "under the hood"?

Thanks

 

I assume we are talking Bachmann - If it is a 32-200 or later it will be DCC 8 pin socket on board. (If it is 32-200DC, BR black 9759, it should have a factory fitted decoder)

IIRC the 8750 was the first Bachmann pannier tank with solid chassis & a DCC socket, then the 57XX got one, prior to that (31-900 series) it was a split chassis and no decoder socket.

 

Keith

Edited September 18, 2018 by melmerby

[Harlequin]

Hi Vic,

 

If you're unsure of what you're doing you could take the body off the loco (carefully), take a photo of the chassis showing the electrical parts and post it here.

 

Then we'll be able to tell you exactly what you've got!

[WIMorrison]

If you want to know whether it has a decoder stick it into the programming track and read CV1.

 

This will tell you if it has a decoder as you get a read back of the address, plus this will give you the address you need to make the loco run.

 

And if you don’t know how to do this then it will be a good start to learning one of the fundamental lessons in DCC as you need to be able to programme the loco address on all locos - otherwise they will all be at the defaul 3 and that will make a terrible mess when they all run at the same time

[melmerby]

If you want to know whether it has a decoder stick it into the programming track and read CV1.

 

 

As long as the decoder is OK, (even then not a guarantee of a fully working decoder, just the address)

 

Keith

[VicZA]

I will open it up a little later and put some photos up as suggested ... until later